Notion launches agents for data analysis and task automation

During the recent 'Make with Notion' event, the productivity platform introduced its groundbreaking AI agent designed to revolutionize data analysis and task automation. This innovative tool will utilize a user's Notion pages and databases to automatically generate insightful notes, meeting analyses, competitor evaluations, and feedback pages. The AI agent boasts the capability to create and update pages and databases with fresh data and new properties. Users can also link external platforms to trigger actions within Notion. For instance, the agent can compile a bug tracking dashboard using information from sources like Slack, email, and Google Drive. Building on the existing Notion AI feature, which previously focused on searching and summarizing content, the new agent can now handle more intricate, multi-step tasks through its agentic AI capabilities. Currently, it can execute a task that lasts up to 20 minutes and spans across hundreds of pages. Users can personalize the agent by setting up a profile page that outlines specific instructions regarding source references, output style, and task updates. Additionally, the agent has a memory function, allowing it to retain key points for future reference, which can be modified by users on the profile page. In demonstration videos, Notion showcased the agent's potential to offer feedback for landing pages, track restaurant information, analyze meeting notes, and generate competitive analysis reports. While these actions currently require manual initiation, the company has indicated that customizable agents capable of operating on a schedule or specific triggers are on the horizon. Notion also plans to introduce a library of templates for agents, providing users with ready-made prompts tailored to various tasks. Over the past two years, Notion has launched several features, including a calendar app, a Gmail client, and an enterprise search tool, all contributing to the foundational elements necessary for creating these automations. Competing platforms like Salesforce, Fireflies, and Read AI have similarly developed their own agents to facilitate information extraction and updates.